PD-1 blockade alone for mismatch repair deficient (dMMR) locally advanced rectal cancer.
Melissa Lumish, Jenna Sinopoli, Zsofia K. Stadler, Jill Weiss, Michelle F. Lamendola-Essel, Rona Yaeger, Neil H. Segal, Imane H. El Dika, Leonard B. Saltz, Marina Shcherba, Ryan Sugarman, Avni M. Desai, J. Joshua Smith, Maria Widmar, Emmanouil P. Pappou, Philip B. Paty, Julio García‐Aguilar, Martin R. Weiser, Luis A. Díaz, Andrea Cercek
Abstract
16 Background: Total neoadjuvant therapy with induction chemotherapy and chemoradiation (chemoRT) is the standard treatment for locally advanced rectal adenocarcinomas. Mismatch repair deficient (dMMR) rectal tumors respond poorly to neoadjuvant chemotherapy. PD-1 blockade is effective in patients with metastatic dMMR colorectal cancers, but its efficacy has not been established in the neoadjuvant setting. The purpose of this study is to evaluate the clinical benefit of neoadjuvant PD-1 blockade in dMMR locally advanced rectal cancer. Methods: We designed a prospective, single-arm, phase II study in which patients with stage II and III dMMR rectal cancer receive neoadjuvant dostarlimab (anti-PD-1) for a total of 6 months. The co-primary objectives are to determine the overall response rate (ORR) and pathologic complete response (pCR) or clinical complete response rate (cCR) with or without chemoRT. Tumor assessment with endoscopic evaluation is performed at baseline, 6 weeks, 3 months and 6 months; imaging is performed at pretreatment baseline, 3 months and 6 months. Patients with cCR by previously established criteria are eligible for non-operative management without chemoRT. Those with residual disease after neoadjuvant dostarlimab receive standard chemoRT. Following chemoRT, any patient failing to achieve a cCR is then managed surgically. Results: A total of 13 patients have been enrolled, with median age 52 years (range 26-78), 77% female, and 92% with node-positive disease by rectal MRI. The ORR is 100% in the 12 patients who have undergone at least a 3-month evaluation. Seven patients have completed induction therapy and all 7 (100%) have achieved a cCR and are undergoing observation without chemoRT or surgery. The rate of progressive disease thus far is 0%. No patients have required chemoRT or surgery. There have been no serious adverse events. Conclusions: Single agent neoadjuvant PD-1 blockade with dostarlimab is effective and well-tolerated in locally advanced dMMR rectal adenocarcinoma and allows patients to avoid chemoradiation and surgery. This suggests a potential new paradigm for treatment of dMMR locally advanced rectal cancer. Follow up and further patient accrual is ongoing. Clinical trial information: NCT04165772.